Can we use a single MLOps pipeline for multiple ML applications?

Course Q&A Machine Learning Engineering for Production(MLOps) Introduction to Machine Learning in Production
1

I would like to know if we can use a single MLOPs pipeline for multiple ML applications or do we require a separate pipeline for each applications.
If it is possible to use a single pipeline could you please specify and give more information regarding it.

1 Like
2

Theoretically you could use part of the same pipeline for different applications, for eg. data preprocessing, a model with several outputs based on same/similar data, api’s for inference can be shared based on demand.

Basically the point is if that part of the pipeline can be used from multiple services (whatever they are and because they require the same tasks to be done) the capacity can be shared within limitations of computing power and availability.